Segment Congruence Analysis Via Information Theory
Jamshid C. Hosseini, Robert R. Harmon, Martin Zwick · PDXScholar (Portland State University) · 1986
The issue of segment interaction arises when multiple segmentation variables are utilized in defining target markets. This paper utilizes the information theoretic measure of transmission to address five issues in overlapping market segments: 1) The existence and the degree of mutual association among a set of segmentation variables. 2) The identification of the segmentation variable (or set of variables) with the highest contribution toward the mutual association. 3) The identification of the segmentation variable with the highest contribution to the reduction of uncertainty in the system of segmentation variables. 4) The predictability of each segmentation variable from a set of exogenous variables (a method to define one or more "distinguished" segmentation variables). 5) The ordering (prioritization) of the exogenous variables based on their contribution to the removal of uncertainty from the distinguished" segmentation variables. Finally, some comparisons are made between the information theoretic approach and the log-linear models used by Green and Carmone (1977). Keywords: Information theory, Transmission, Mutual information, Reconstructability Analysis, Market segmentation, Segment congruence analysis